scholarly journals Multiple testing correction over contrasts for brain imaging

2019 ◽  
Author(s):  
Bianca A. V. Alberton ◽  
Thomas E. Nichols ◽  
Humberto R. Gamba ◽  
Anderson M. Winkler
Keyword(s):  
Brain Imaging ◽  
Linear Model ◽  
Multiple Testing ◽  
Image Representation ◽  
General Linear Model ◽  
Parameter Estimates ◽  
Multiple Testing Correction ◽  
Testing Problem ◽  
Multiple Testing Problem ◽  
The Brain

AbstractThe multiple testing problem arises not only when there are many voxels or vertices in an image representation of the brain, but also when multiple contrasts of parameter estimates (that is, hypotheses) are tested in the same general linear model. Here we argue that a correction for this multiplicity must be performed to avoid excess of false positives. Various methods have been proposed in the literature, but few have been applied to brain imaging. Here we discuss and compare different methods to make such correction in different scenarios, showing that one classical and well known method is invalid, and argue that permutation is the best option to perform such correction due to its exactness and flexibility to handle a variety of common imaging situations.

Cortical and subcortical morphology deficits in cerebral gray matter in patients with schizophrenia and not affected siblings

2016 ◽  
Vol 33 (S1) ◽  
pp. s249-s249
Author(s):  
F. Pastoriza ◽  
L. Galindo ◽  
A. Mané ◽  
D. Bergé ◽  
N. Pujol ◽  
...  
Keyword(s):  
Magnetic Resonance ◽  
Linear Model ◽  
Cortical Thickness ◽  
A Priori ◽  
Multiple Comparisons ◽  
General Linear Model ◽  
General Linear ◽  
Group Differences ◽  
Competing Interest ◽  
The Brain

ObjectiveExplore the basis of cortical morphometry in patients with schizophrenia and non-affected siblings by Magnetic Resonance Structural analyzing cortical thickness.MethodsTwenty-nine patients with schizophrenia treated with atypical antipsychotics and clinically stable in the last 6 months were recruited. Twenty-three not affected siblings of patients with schizophrenia and 37 healthy volunteers were recruited. Magnetic Resonance Structural was performed. FreeSurfer the brain imaging software package for analysis of Cortical Thickness is used. In the analysis of group differences in cortical thickness (CT) with the general linear model (GLM), the P-value was established in 0003 following the Bonferroni correction to control for multiple comparisons (seven regions of interest a priori in each hemisphere).ResultsSignificant differences in cortical thickness between patients and healthy controls. Differences between groups were calculated by general linear model (GLM) with age and sex as covairables (Table 1).ConclusionsIn applying the correction for multiple comparisons, differences in bilateral-lateral orbitofrontal, medial orbitofrontal-right and left temporal transverse frontal cortex are significant. Our study replicates previous findings and provides further evidence of abnormalities in the cerebral cortex, particularly in the frontal and temporal regions, being characteristic of schizophrenia.Disclosure of interestThe authors have not supplied their declaration of competing interest.AcknowledgementsL. Galindo is a Rio-Hortega-fellowship-(ISC-III; CM14/00111).

scholarly journals Commentary on Vul et al.'s (2009) “Puzzlingly High Correlations in fMRI Studies of Emotion, Personality, and Social Cognition”

2009 ◽  
Vol 4 (3) ◽  
pp. 291-293 ◽  
Author(s):  
Thomas E. Nichols ◽  
Jean-Baptist Poline
Keyword(s):  
Social Cognition ◽  
Multiple Testing ◽  
Current Practice ◽  
Testing Problem ◽  
Multiple Testing Problem ◽  
Inference Methods ◽  
Post Hoc

The article “Puzzlingly High Correlations in fMRI Studies of Emotion, Personality, and Social Cognition” ( Vul, Harris, Winkielman, & Pashler, 2009 , this issue) makes a broad case that current practice in neuroimaging methodology is deficient. Vul et al. go so far as to demand that authors retract or restate results, which we find wrongly casts suspicion on the confirmatory inference methods that form the foundation of neuroimaging statistics. We contend the authors' argument is overstated and that their work can be distilled down to two points already familiar to the neuroimaging community: that the multiple testing problem must be accounted for, and that reporting of methods and results should be improved. We also illuminate their concerns with standard statistical concepts such as the distinction between estimation and inference and between confirmatory and post hoc inferences, which makes their findings less puzzling.

scholarly journals An experimentally-derived measure of inter-replicate variation in reference samples: the same-same permutation methodology

2019 ◽  
Author(s):  
David C. Handler ◽  
Paul A. Haynes
Keyword(s):  
Multiple Testing ◽  
Large Scale ◽  
Reference Sample ◽  
Q Value ◽  
Testing Problem ◽  
False Discovery ◽  
Proteomics Experiment ◽  
Multiple Testing Problem ◽  
High Throughput Data Analysis ◽  
Reference Samples

AbstractThe multiple testing problem is a well-known statistical stumbling block in high-throughput data analysis, where large scale repetition of statistical methods introduces unwanted noise into the results. While approaches exist to overcome the multiple testing problem, these methods focus on theoretical statistical clarification rather than incorporating experimentally-derived measures to ensure appropriately tailored analysis parameters. Here, we introduce a method for estimating inter-replicate variability in reference samples for a quantitative proteomics experiment using permutation analysis. This can function as a modulator to multiple testing corrections such as the Benjamini-Hochberg ordered Q value test. We refer to this as a ‘same-same’ analysis, since this method incorporates the use of six biological replicates of the reference sample and determines, through non-redundant triplet pairwise comparisons, the level of quantitative noise inherent within the system. The method can be used to produce an experiment-specific Q value cut-off that achieves a specified false discovery rate at the quantitation level, such as 1%. The same-same method is applicable to any experimental set that incorporates six replicates of a reference sample. To facilitate access to this approach, we have developed a same-same analysis R module that is freely available and ready to use via the internet.

scholarly journals The multiple testing problem for Box-Pierce statistics

2014 ◽  
Vol 8 (1) ◽  
pp. 497-522 ◽  
Author(s):  
Tucker McElroy ◽  
Brian Monsell
Keyword(s):  
Multiple Testing ◽  
Testing Problem ◽  
Multiple Testing Problem
Sign in / Sign up

Export Citation Format

Share Document